[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: scaling

From   Maarten buis <>
Subject   Re: st: scaling
Date   Sat, 17 Nov 2007 22:51:50 +0000 (GMT)

--- Renuka Metcalfe <> wrote:
> I am using a cross-section data. I am looking at the
> impact of "x" on "Y", runing an OLS. I ran something
> like
> .regress Y x1, x2, x3
> Some of my coefficients are very large. For example,
> 648.
> I cannot log the dependent variable, since some of the
> some workplaces had minus "Y".
> I guess I could divide "Y" by 1000. But I not certain
> if this would be correct.

dividing by 1000 is certainly less dramatic than taking
the log. Actually you can say you change nothing at all
except that you get an model that is easier to estimate 
and easier to interpret. 

Say your dependent variable is income in euros. If you 
divide by a 1000 you get income in 1000s of euros. 
Whether you say that an extra level of education leads 
to 500 euros more income or .5 thousand euros income 
makes no difference, it is the same thing.

Hope this helps,

Maarten L. Buis
Department of Social Research Methodology
Vrije Universiteit Amsterdam
Boelelaan 1081
1081 HV Amsterdam
The Netherlands

visiting address:
Buitenveldertselaan 3 (Metropolitan), room Z434

+31 20 5986715

Yahoo! Answers - Got a question? Someone out there knows the answer. Try it
*   For searches and help try:

© Copyright 1996–2017 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index